Re: [Nbd] Question about the expected behaviour of nbd-server for async ops

[Alex Bligh <alex@...872...>]

Wouter,

--On 29 May 2011 11:21:47 +0200 Wouter Verhelst <w@...112...> wrote:

> That might be interesting, but then any protocol extensions would
> require IETF updates. I'm not sure that's ideal. Currently nbd is a
> Linux-internal protocol, and we can change it at will; we'd lose that
> ability to some extent. Whether that's a fatal problem is, of course, up
> for discussion.
>
> Having said that, I do agree that having a more formal description of
> the protocol is a good thing; proto.txt is a good start, but it's not
> fully there yet.
>
> At any rate, if there's going to be an IETF draft, I'm going to be a
> (co-)author ;-)

I really meant "written up in the style of an IETF draft", rather than
actually submitted as one, i.e. use MUST/MAY/SHOULD conventions, use
IETF formatting etc.

Now you mention it, it might be possible to submit it as an
experimental draft (i.e. not something designed to progress down
standards track yet) without introducing undue friction if we
want to change something.

--
Alex Bligh